Selective targeting of Abeta42 may be an ideal therapeutic strategy for Alzheimer's disease (AD). We Deviously reported that certain NSAIDs modulate Abeta42 production. We have also identified compounds ;hat increase Abeta2. We now refer to these compounds as gamma-secretase modulators (GSMs). The signature activity of GSMs is that they minimally alter total Abeta production but shift the y-secretase cleavage site. Abeta42 lowering GSMs increase shorter Abeta peptides and Abeta42 raising GSMs decrease shorter Abeta peptides. GSMs shift y-cleavagethrough a relatively unprecedented mechanism. Instead of targeting gamma-secretase, they target substrate hence substrate-targeting GSM (stGSM). These substrate-targeting GSMs bind substrate (APP/APP CTF) and product (Abeta) in a region that corresponds to residues 29-36 of Abeta. This region of Abeta is critical for aggregation. We find that stGSMs do in fact nhibit Abeta aggregation, and that many compounds identified as aggregation inhibitors are GSMs. These findings suggest that stGSMs have two potential therapeutic actions, alteration in Abeta42 production and nhibition of Abeta aggregation, that may synergistically reduce Abeta deposition in AD. As shorter Abeta peptides may also be protective, it is possible that GSMs which increase the levels of these shorter peptides may also have a third beneficial therapeutic action. In this proposal we will extend mechanistic studies regarding substrate targeting by GSMs and evaluate linked hypotheses regarding the in vivo mechanism of action of GSMs. We will perform acute and chronic dosing studies with Abeta42 lowering and raising GSMs in APP and BRI-Abeta42 mouse models to evaluate the relative contribution of altering Abeta42 production vs. altering aggregation with respect to effect on Abeta loads and other AD-like pathologies. These in vivo studies will enable us to model possible changes seen in human trials in Project 3. In close collaboration with Project 1, we will also determine if elevations in shorter Abeta peptides are protective. For these later studies we will use recombinant adenoassociated virus (rAAV) to deliver BRI-Abeta fusion constructs encoding short Abeta peptides to the brain of neonatal APP mice. This methodology creates "somatic brain transgenics" and will allow us to rapidly evaluate the effects of shorter Abeta peptides on Abeta deposition. Collectively, these studies should provide additional insight into the mechanism whereby GSMs shift Abeta cleavage and their protective effects in vivo.